Keywords: ecology, New Zealand, palaeoclimatology, pollen, Quaternary, spores

Abstract

Because of the widespread occurrence of organic sediments, New Zealand is ideally suited to the study of Cenozoic vegetation and climates by fossil spores and pollen. Primary descriptions and interpretations have already been made for most parts of the country and these are widely relied upon by a number of other disciplines.

In this paper we review the reasoning by which past vegetation and environments are reconstructed. Recent advances in plant ecology, historical ecology, eco-physiology, and palaeobotany, related to palaeoecology are reviewed.

The paper includes a comprehensive table giving the ecological ranges of plants found in the New Zealand pollen and spore record.

Introduction

Quaternary pollen analysts have long had two complementary roles. The first is the reconstruction of past vegetation using pollen and spores preserved in sediment and unconsolidated sediments ranging from-calcareous loess to peat. The second is the ecologic interpretation of paleovegetation in terms of past environments, usually climate. The many pollen diagrams now available for New Zealand provide primary descriptions of past floras, vegetation and climates during the Tertiary and Quaternary as well as answers to more specific research questions e.g. identification of stadials/interstadials (Moar 1971, 1973; Moar & Suggate 1979; Mildenhall 1980).

Fossil pollen and spores are direct evidence only for past floras, with some indication of abundance (Faegri 1961). The reasoning by which the higher levels of ecological interpretation are reached are complex, difficult to define and usually subjective (cf. Rymer 1976). Difficulties of interpretation may or may not be able to be solved but this depends on the aim of the pollen analyst and the scale of the problem (Oldfield 1970).

The difficulties include:

• 1) limits of specific pollen and spore identification, (often related to preservation).

• 2) distinguishing between proximity and abundance of the source plants.

• 3) differences in the area of vegetation represented by different pollen and spore types, defined as the pollen source area, and related to pollen productivity and dispersal power (Flenley 1973).

• 4) changes over time in the pollen source area with change in floristics and structure of the local vegetation (Tauber 1965, 1967.)

• 5) The conservative nature of many pollen types, especially those produced by rapidly speciating groups; this feature is an important problem in the New Zealand flora.

• 6) The present-day ecology of plants represented in pollen assemblages is still often imperfectly known. The later part of this paper (Table 2) attemps to summarise current knowledge for New Zealand.

In this paper we review knowledge of present-day vegetation/pollen relationships to aid the critical appraisal of paleoecological data from New Zealand. The paper relies extensively on the studies of N.T. Moar and P. Wardle (Botany Division, DSIR) and D. T. Pocknall (N.Z. Geological Survey).

Representativity and Pollen Transport

Representativity is a measure expressing the relationship between species abundance and amount of pollen recorded at a particular location. Although the relationships can be expressed quantitatively for a particular site, (Davis 1963, Flenley 1973, Pocknall 1978, 1979), it is usually impossible to define the precise pollen source area. Moreover, differences in pollen productivity and dispersal power lead to different areas of vegetation being represented by the different pollen and spore texa at any time. The differences and variations between sites and over time have therefore led to representativity values being express qualitatively:

• i Over-represented: pollen percentage is greater than the percentage of the source plants within the vegetation.

• ii Well-represented: pollen percentage is approximately equal to percentage of the source plants within the vegetation.

• iii Under-represented: pollen percentage is markedly smaller than percentage of the source plants within the vegetation, or zero.

These classes are based on the experience of the pollen analyst but are now usually supported by studies of the modern pollen rain in the same area as the fossil study. New Zealand examples include Moar (1969, 1970, 1971), Wardle (1971a), McKellar (1973), Myers (1973), Mildenhall (1976), Dobson (1976) and particularly Pocknall (1978, 1979, 1980).

Few texa indigenous to New Zealand have intrinsically high powers of pollen dispersion: Cyathea smithii-type, Dacrydium cupressinum, Nothofagus fusca type, Coprosma and Gramineae.

Other plants have lesser powers of pollen dispersal but nevertheless some 25 pollen and spore types are frequently encountered due to the widespread occurrence of the source plants across New Zealand:

Compositae, Phyllocladus, Podocarpus ferrugineus, P. spicatus, P. totara-type, Dacrydium bidwillii/biforme, Dacrydium colensoi, Nothofagus menziesii, Quintinia, Ascarina, Metrosideros, Weinmannia, Elaeocarpus, page 39 Griselinia, Pseudopanax, Myrsine, Pennantia, Leptospermum, Dicksonia, Cyathea dealbata-type, Phymatodes, Histiopteris, Monolete fern spores, Lycopodium spp. and Dacrycarpus dacrydioides. These types are usually the basis for interpreting and comparing pollen sequences even although they usually form only a small part of the total pollen flora preserved at any site. Consquently, there is often a broad similarity between fossil pollen diagrams from widely separated and floristically distinct areas in New Zealand.

The majority of New Zealand pollen and spore types are seldom found in modern assemblages unless the source plants are close to the site or water transport is involved. That is, these taxa are severely under-representative of the non-local vegetation but may be greatly over-represented if the plants grew at the pollen site. Most insect-pollinated trees, and the great majority of shrubs and herbs fall into this category. Wind-pollination is not necessarily a guarantee of good representation, e.g. Cyperaceae and Dacrycarpus.

A small group of plants exists for which pollen is virtually never found for instance the important hardwood trees Beilschmiedia tarairi and B. tawa (Macphail 1980). Such non-represented species are a “blind spot in New Zealand vegetation history, and pollen assemblages from Beilschmiedia- dominated forest are entirely misleading due to over-representation of local shrubs and distant strong pollen sources, usually conifers. Conversely severe under-representation can be an advantage in that even very low percentages of these taxa is good evidence for the local presence of the source, and for particular local environments if the source, plants have a well-defined ecologic range, such as Dactylanthus (Macphail and Mildenhall, 1980).

In addition to the intrinsic dispersal properties of individual pollen and spores types, the mode of pollen transport to the site also influences the magnitude of pollen deposition and size of the pollen source area. Studies in North-West Europe (Tauber 1965, 1967) suggest that in forest, pollen is transported to most sites in streams and surface run-off (C_w), through the spaces between tree trunks (C_t), above the canopy by wind (C_c) and by rain drops (C_r). Jacobson and Bradshaw (1981) have defined an additional pollen component (C_g) for pollen deposited vertically downwards from overhanging the sampling site. Tauber's model predicts that the relative proportions of these pollen components at a site is determined by the floristic and structural characteristics of the surrounding vegetation and nature of the site, particularly size and presence or not of inflowing streams. A corollary is that a change in site character, e.g. from a lake to a sedgeland, will alter the pollen source area and at such times, changes in vegetation seen via pollen analysis may have nothing to do with climate, (Ladd 1979).

Because spores and pollen are abundantly preserved in water-logged sediments such as peats and lake muds, basins where these sediments can accumulate continuously over long periods of time are favoured for pollen analytical studies. Fig. 1 shows the generalized relationship between the size of a site that has no inflowing stream and the relative proportions of pollen originating from different areas around the site: local- from plants growing within 20m of the basin edge; extra-local- from plants growing between ca. 20-300m from the basin edge; regional- growing at distances greater than ca. 300m (Jacobson and Bradshaw 1981 p. 82; Fig. 1). For shrub-covered peat bogs and basins of less than 1 ha, within forest, the major source of pollen will be local (C_g, C_w and C_t) Regionally derived page 40

Fig. 1 The generalized relationship between the size of a site that has no inflowing stream, the relative proportions of pollen originating from different areas around the site and level of vegetation reconstruction (after Jacobson and Bradshaw 1981)

pollen will penetrate into the sampling site (probably as C_r) but this long distance transported component will be dominated by locally derived types. The model predicts that as the dominated by locally dervied types. The model predicts that as the canopy opens up^*, or the sampling site becomes incresingly distant from the dominant stratum, so extra-local and regional elements become increasingly well represented. Sub-tropical and tropical rainforest may however be an exception (Kershaw and Hyland 1975). Because pollen production of these rainforest trees is very low, domination of assemblages by local pollen does not occur and pollen deposition appears to be an accurate reflection of the regional vegetation.

All undisturbed hardwood, podocarp and beech forests and most forms of scrub in New Zealand are evergreen and closed (sensu Specht 1970). Most of the lakes utilized for paleovegetational research are small (less than 100m diameter) or if larger (Pocknall 1980) sampled close to the shore. Consequently, if the site is below the climatic treeline, the pollen sequences will preferentially record shifts in dominance with the local canopy and riparian communities. These shifts will be accompanied by some less well-defined indications of shifts in dominance elsewhere, particularly if the C_w component is important. So far there is no way to distinguish between a few strong pollen producers near the sampling site and, abundance of the same taxa in the regional vegetation. Consequently patterns in vegetation at the local and extra-local scale can only be detected by utilizing suites of sites dominated by local pollen.

The reverse situation occurs in low, open vegetation such as heath, page 41 grassland and herbfield. Local pollen influx may or may not be low but is usually augmented by pollen from the regional vegetation (C_r, C_c) and, at sites above the treeline, dominated by these taxa (Table 1). Here, occurrences of pollen from severely under-represented taxa that are restricted to cold climates, e.g. Donatia novae-zelandiae, Notothlapsi and Phyllachne may be the only definite indicators of the local enviroment whilst part of the long distance transported pollen is derived from the prevailing winds, much is derived from a preferential upslope transport of forest pollen types from lower elevations (Moar 1970, Heusser 1974, Macphail 1975; 1979, Hope 1976, Salgado-Labouriau 1979). With over-representation of distant, strong pollen dispersers, the boundaries between different vegetation types are blurred in the pollen record, particularly the precise location of the climatic treeline relative to the sampling site.

Modern pollen data supporting the above generalisatons are given in Table 1. These show the broad similarity between pollen assemblages from within Nothofagus temperate rainforest (Lake Hanlan, Shag Tarn) and those from sites in low, open vegetation (Denniston) or above the climatic treeline (Saddle Lakes). At Saddle Lakes Dacrydium bidwillii-type, Donatia novea zelandiae and Phyllachne are the only pollen evidence for the presence of the local alpine vegetation. The high percentage of Gramineae pollen at Denniston may reflect the local grassland but are more likely to be due to the occurrence of a few plants in the tank. The influence of close vegetation on pollen assemblages is more clearly seen in the moss samples from Lake Hanlan and Shag Tarn. At Lake Hanlan, the forest floor is virtually devoid of all but cryptogams and the moss sample is overwhelmingly dominated by Nothofagus fusca-type pollen. Despite the proximity of several emergent Dacrydium cupressinum trees, podocarp pollen is less that 5% implying very low wind velocities within the forest stand. At Shag Tarn, a moderately dense substratum of Coprosma spp. occurred and this is reflected in the relatively high shrub pollen percentages. Conversely the surface lake muds at both sites contained diverse palynofloras representing a range of local (riparian) and upland plant communities within the catchments.

Reconstruction of Paleovegetation

Knowledge of the relative pollen dispersal class of an individual pollen type, and identification of its probable mode of transfer to the sampling site, enable the pollen analyst to assess the relative abundance of that taxon in a fossil flora. Reconstruction of paleovegetation requires a second step, identification of the fossil plant communities.

A plant community can be defined as an assemblage of plants living together in an area (of any size) during the same time and in relatively stable proportions (Atkinson et al. 1968). Since fossil pollen sequences are an incomplete record of past floras and usually integrate the pollen production from a number of different communities, it is difficult directly to identify past plant communities. Instead one or more of three indirect approaches are adopted (Birks and Birks 1980):

1. Using statistical techniques to establish recurrent groups of fossil taxa, i.e. assemblages of fossil pollen and spores consistently occurring together, or numerically correlated in a series of samples within (and between) pollen sequences.

2. Applying to past floras the ecological and phytosociological preferences of modern taxa.

If a species can be shown to occur in a limited number of associations, verifiable by objective ecological techniques, then the pollen of that species is assumed to indicate that association in the past. A related assumption is that the structure of the paleovegetation can be deduced by analogy when the past associations appear to agree with present-day well-defined associations.

3. Matching fossil pollen assemblages with modern assemblages from known vegetation types (the “finger-printing approach, Martin and Gray 1962, Wright 1967).

Approaches 2 and 3 depend on finding modern analogues so that the present can be extrapolated into the past. Both are common palynologic practice, at least implicitly, but are misleading if applied without caution to early Quaternary or pre-Quaternary assemblages. Their successful use requires sound knowledge of modern ecological preferences.

Another problem is the difficulty of indentifying stable plant communities that are independent of climate, such as edaphically controlled Dacrycarpus semi-swamp forest and Westland “pakihi sedgelands of gley podsols (Rigg, 1962). Both communities include well-represented and palynologically distinct taxa.

Plant communities above the climatic treeline in New Zealand are usually well-defined floristically and uniform throughout the country. The tallest stratum is dominated by combinations of Compositae, Gramineae, Umbelliferae, Rubiaceae, Scrophulariaceae, Epacridacae and Liliaceate but structural and floristic analogues can also occur in exposed or semi-arid areas at lower elevations. Communities below the treeline are less easily defined. Although there are relatively few canopy species, mostly Podocarpaceae, Nothofagus spp., Weinmannia, Beilschmiedia, Elaeocarpus, Metrosideros and Quintinia. There is also a gradual replacement of mixed podocarp-hardwood stands by Nothofagus with increasing altitude, and latitude. Consequently these taxa can occur in any combination or proportion within the same climatic regime, (Holloway 1954, Robbins 1962, Nicholls 1976, Park and Walls 1978). Podocarp stands are widely recognized as having unstable population structures under present-day climates (Herbert 1980) and in some localities they are being competitively excluded by hardwoods and Nothofagus (Wardle 1963, Clayton-Green 1977).

Accordingly, Wardle (1964) has questioned the existence of stable forest associations and concluded that the component species have overlapping but largely independent distributions. These distributions are determined not only by differences in reproductive biology and habitat requirements (determining rates of migration and performance within the forest succession) but also by past events. The latter include the severe but geographically variable effects of Quaternary glaciation (Moar 1973, Burrows 1979) and volcanism (McKelvey 1963, McGlone and Topping 1977) as well as perennial phenomena such as landslipping, catastrophic weather and fires (Molloy 1969, Wardle 1978).

If canopy species have behaved individualistically through time, then it is clearly unwise to press analogies based on the modern vegetation too closely when inferring past associations. For example, Walker and Flenley (1979) have demonstrated that whilst some taxa in New Guinean ‘podocarp-hardwood-Nothofagus’ rainforest occur together with great fidelity through a variety of environmental vicissitudes, other taxa associated with them may behave independently when viewed on longer time scales during the late Quaternary. In New Zealand, this problem has to a large extent been avoided by interpreting past vegetation in terms of formations defined by pollen dominants and using minor, less well-represented taxa to suggest thermal or geographic variatons, (Moar 1971, 1973, McGlone and Topping 1973, 1977, Dobson 1978). This approach is appropriate to developing regional paleovegetational reconstructions but smaller scales of pattern have mostly yet to be interpreted.

Reconstruction of Past Environments

Reconstruction of past environments is usually based on both abiotic (lithology, geochemistry, site geomorphology, isotopes etc.) and biotic evidence, of which fossil pollen data is but one part. The most common method in pollen analysis is an extension of the ‘indicator species’ approach, the application backwards in time of known preference of (i) individual taxa with well-defined narrow environmental ranges or (ii) plant communities of taxa having similar ranges. Alternatively, statistical methods which relate pollen values to one or more environmental parameters using discriminant functions or regression equations may be used. A New Zealand example of discriminant function analysis relating groups of pollen and spores to temperature is provided by Harris et al. (1976). Transfer regression equations have been successfully used in a rainforest area of Chile by Heusser and Streeter (1980) and may be found applicable to New Zealand once the data for modern pollen spectra are adequate.

Much of the present information concerning the environmental tolerances of phytosociologically important species in New Zealand comes from compilations of available herbarium and field records: Ascarina lucida (McGlone and Moar 1977), Chionochloa rigida(Mark 1965), Dacrydium cupressinum (Franklin 1968), Leptospermum spp. (Burrell 1965), Metrosideros umbellata (Wardle 1971b), Nothofagus menziessii (Wardle 1967), N. solandri (Wardle 1970), Olearia colensoi (Wardle et al. 1970), Phyllocladus alpinus (Wardle 1969) and Weimannia racemosa (Wardle 1966). Agathis australis (Eckroyd 1982), Beilschmiedia tawa (Knowles and Beveridge 1982). Not all these taxa can be specifically identified by pollen. Ecological data on other equally important taxa is scattered and often unpublished.

Nevertheless, the information available does indicate that the majority of, if not all, phytosiciologically important trees have very wide environmental tolerances. In particular the taller podocarps are able to survive conditions which prevent regeneration by seed but, because of their longevity, are still able to influence the stability of rainforest communities over this period. Hardwood taxa appear to have shorter life-spans and would be excluded more rapidly under unfavourable conditions. However, the pollen of neither group provides a sensitive indication of climate except over long periods of time, suggested to be in excess of two thousand years.

Subcanopy trees and shrubs with more restricted distributions but with good representation are proving more rewarding as climatic indicators Ascarina lucida is one example (McGlone and Moar 1977) and trends in this taxon may serve to correlate early to middle Holocene pollen sequences across New Zealand. Dodonaea may be another (W. F. Harris pers. comm.). Because many of these taxa are under-represented, it is necessary to count large numbers of pollen or scan pollen slides for trace occurrences combined with a lower pollen count.

A remarkable example of the use of experimental plant physiology to interpret early postglacial sequences in the eastern South Island is given by Wardle and Campbell (1976) in their studies on frost-tolerance of plants found as pollen.

Table 2 lists pollen and spore types that (i) occur regularly and in some abundance in late Quaternary sediments or (ii) are rare but specifically identifiable or (iii) indicative of particular environments. Relative pollen dispersal classes are based on Dobson (1976), Pocknall (1978, 1979, 1980) and Macphail (unpub. results): Comments on the distribution and ecology take into account the available ecological literature (Wardle 1973, Orwin 1974) and extensive field observations by Wardle (1975, 1977) and D. R. McQueen but cannot take account of many anomalous distributions (Holloway 1954). Plant names and climate zones follow Allan (1961), Moore and Edgar 1970) and Wardle (1964).

Conclusions

There can be no doubt that pollen analysis has been spectacularly successful in providing primary descriptions of past vegetation and environments in New Zealand, particularly for the Late Quaternary period. Published descriptions are lacking only from North Auckland (M. S. McGlone pers. comm.) but all parts of the country will benefit from additional studies. At this scale trends in vegetation structure and climate parallel those in South-eastern Australia and Southern Chile (Macphail 1979, Heusser 1974).

Rather less is known of patterns in vegetation and climate on the smaller spatial and temporal scale. Where specific values for postglacial climates have been determined from fossil pollen assemblages, these are supported by macrofossils, e.g. Lintott and Burrows (1973), Soons and Burrows (1978), McGlone et al. 1978).

Whether New Zealand rainforest comprises independently distributed species as suggested by Wardle (1964) or whether stable associations and unifying ecological trends exist, it is clear that even a single cycle of regeneration or establishment is too long for any individual to observe directly. Processes in forest succession may be inferred by integrating observations made on suites of stands, as Wardle (1980) has done using moraines in the Westland National park. Alternatively, well-dated long pollen sequences can provide detailed evidence for successional trends in the one place, (Macphail, in prep.) Additional insights are being provided by historical records (Beever 1981). Macrofossil assemblages (e.g. Molloy and Cox 1972, Campbell et al. 1973, Drake and Burrows 1980) as well as plant population studies (e.g. Clayton-Green 1977, Wardle 1978) and tree nutritional data (Adams 1976), have also made significant contributions to our knowledge of successions past and present.

Many fundamental questions remain unaddressed. What is certain is that New Zealand's abundant, deep and varied organic deposits will enable the specific paleoecological queries to be isolated and answered, be they botanic, edaphic or climatic.

Acknowledgements

We are indebted to many for their help in the production of this paper, which arises from work during a Post-Doctoral Fellowship help by M. K. Macphail at the Botany Department, Victoria University of Wellington. Terese Hughes (University of Tasmania) and Maureen Cooper (Victoria University of Wellington) for their considerate and considerable typing efforts. D. C. Mildenhall and D. T. Pocknall (N.Z. Geological Survey), and N. T. Moar and M. S. McGlone (Botany Division, DSIR) aided by critically assessing the manuscript.

References

Adams, J. A. 1976: Nutrient requirments of four Nothofagus species in North Westland, as shown by foliar analysis. N.Z. Journal of Botany 14: 211-23.

Allan, H. H. 1961: Flora of New Zealand Vol. 1. Government Printer, Wellington. 1085pp.

Atkinson, I. A. E.; Jenkins, P. F.; Druce, A. P. 1968: Definitions and discussions of some concepts and terms relating to terrestrial ecosystems. Tuatara 16:98-110.

Beever, J. 1981: A map of pre-European vegetation of lower Northland, New Zealand. N.Z. journal of Botany 19: 105-10.

Birks, H. J. B.; Birks, H. H. 1980: “Quaternary Paleoecology. Arnold, London. 283pp.

Birks, H. J. B.; Webb, T.; Berti, A. A. 1975. Numerical analysis of surface samples from central Canda: a comparison of methods. Paleobotany and Palynology 20: 133-69.

Bowman, D. M. J. S.; Jackson, W. D. 1981. Vegetation succession in Southwest Tasmania. Search 12: 358-62.

Burrell, J. 1965: Ecology of Leptospermum in Otago. N.Z. Journal of Botany 3: 3-16.

Burrows, C. J. 1979: A chronology for cool-climate episodes in the Southern Hemisphere 12000-1000 Yr. B.P. Paleogeography, paleoclimatology, paleoecology 27: 287-347.

Cameron, R. J. 1954: Mosaic or cyclic regeneration in North Island podocarp forests N.Z. Journal of Foresty 7: 55-65.

Campbell, E. O.; Heine, J. C.; Pullar, W. A.; 1973. Identification of plant fragments and pollen from peat deposits in Rangitaiki Plains and Maketu Basins. N.Z. Journal of Botany 11: 317-30.

Clayton-Greene, K. 1977: Structure and origin of Libocedrus bidwillii stands in the Waikato District, New Zealand. N.Z. Journal of Botany 15: 19-28.

Close R. C.: Moar, N. T.; Tomlinson, E. I.; Lowe, A. D.; 1978. Aerial disperal of biological material from Australia to New Zealand. International Journal of Biometerology 22: 1-19.

Davis, M. B. 1963: On the theory of pollen analysis. American Journal of Science 216: 897-912.

Dobson, J. R. 1976: Modern pollen spectra from Chatham Island, New Zealand. N.Z. Journal of Botany. 14: 341-7.

—— 1978: A vegetation history from north-east Nelson, New Zealand. N.Z. Journal of Botany. 16: 371-8.

Drake, H.: Burrows, C. J. 1980. The influx of potential macrofossils into Lady Lake, north Westland, New Zealand. N.Z. Journal of Botany 17: 257-74.

Druce, A. P. 1961: Mountain vegetation of the North Island. N.Z. Soil News 3: 95-107.

Eckroyd, C. E. 1982: Biological flora of New Zealand: Agathis australis (D. Don) Lindl. (Araucariaceae) Kauri N.Z. Journal of Botany 20:17-27.

Faegri, K. 1966: Some problems of representativity in pollen analysis. Paleobotanist 15: 135-40.

Flenley, J. R. 1973: The use of modern pollen rain samples in the study of vegetational history of tropical regions. In Birks, H. J. B. and West, R. G. (Eds) “Quaternary Plant Ecology . Blackwell, Oxford, pp. 131-41.

Franklin, D. A. 1968: Biological flora of New Zealand 3. Dacrydium cupressinum Lamb (Podocarpaceae) Rimu. N.Z. Journal of Botany 6: 493-513.

Harris, W. F.: Norris, G. 1972. Ecological significance of recurrent groups of pollen and spores in Quaternary sequences from New Zealand. Paleogeography, Paleoclimatology, Paleoecology 11: 107-24.

Harris, W. F.; Darwin, J. H.; Newman, M. J. 1976: Species clustering and New Zealand Quaternary climate. Paleogeography, Paleoclimatology, Paleoecology. 19: 231-48

Herbert, J. 1980: Structure and growth of dense podocarp forests at Tihoi, Central North Island and the impact of selective logging. N.Z. Journal of Forestry 25,1: 44-57.

Heusser, C. J. 1974: Vegetation and climate of the southern Chilean Lake District during and since the last interglaciation. Quaternary Research 4: 290-315.

——; Streeter, S. S. 1980: A temperature and precipitation record of the past 16,000 years in Southern Chile. Science 210: 1345-7.

Holloway, J. T. 1954: Forest and climates in the South Island of New Zealand. Transactions of the Royal Society of New Zealand 82: 329-410.

Hope, G. S. 1976: The vegetational history of Mt. Wilhelm, Papua New Guinea. Journal of Ecology 64: 627-61.

Jackson, W. D. 1968: Fire, air, water and earth - an elemental ecology of Tasmania. Proceedings of the ecological Society of Australia 3: 9-16.

Jacobson, G L.; Bradshaw, R. H. W. 1981. The selection of sites for paleovegetation studies. Quaternarr Research 16: 80-96.

Johnson, P. N. 1972; Applied ecological studies of off-shoreline vegetation at Lakes Te Anau and Manapouri. Proceedings N.Z. Ecological Society 19: 120-142.

Kershaw, A. P.; Hyland, B. P. M. 1975. Pollen transfer and periodicity in a rainforest situation. Review of Paleobotany and Palynology 19: 129-38.

Knowles, B.; Beveridge, A. E. 1982: Biological flora of New Zealand. 9. Beilschmiedia tawa (A. Cunn) Benth et Hook. f. ex Kirk (Lauraceae) tawa. N.Z. Journal of Botany 20,1: 37-54.

Ladd, P. G. 1979: A short pollen diagram form rainforest in highland eastern Victoria. Australian Journal of Ecology 4: 299-37.

Lintott, W. H.; Burrows, C. J. 1973: A pollen diagram and macrofosils from Kettlehole Bog, Cass, South Island, New Zealand. N.Z. Journal of Botany, 11: 269-82.

Mcglone, M. S.; Moar, N.T. 1977: The Ascarinadecline and postglacial climatic change in New Zealand. N.Z. Journal of Botany 15: 485-9.

——; Nelson, C. S.; Hume, T. M. 1978: Palynology, age and environmental significance of some peat beds in the upper Pleistocene Hinuera Formation, South Auckland, New Zealand. Journal of the Royal Society of New Zeland 8: 385-93.

——; Topping, W. W. 1973: Late Otiran/early Aranuian vegetation in the Tongariro area, central North island, New Zealand. N.Z. Journal of Botany 11: 283-90.

——; Topping, W. W. 1977: Aranuian (post-glacial) pollen diagrams from the Tongariro region, North Island, New Zealand. N.Z. Journal of Botany 15,4: 749-760.

McKellar, M. H. 1973: Dispersal of Nothofaguspollen in eastern Otago, South Island, New Zealand. N.Z. Journal of Botany 11: 305-10.

McKelvey, P. J. 1963: The synecology of the West Taupo indigenous forest. N.Z. Forest Service Bulletin 14: 1-127.

Macphail, M. K. 1975: Late Pleistocene environments in Tasmania. Search 6: 259-300.

—— 1979: Vegetation and climates in Southern Tasmania since the last glaciation. Quaternary Research 11: 306-41.

—— 1980: Fossil and modern Beilschmiedia pollen in New Zealand. N.Z. Journal of Botany 18: 453-7.

——; Mildenhall, D. S. 1980: Dactylanthus taylorii: in, North-West Nelson? N.Z. Journal of Botany 18:149-52.

Mark, A. F. 1965: Effect of management practices on narrowleaved snow tussock, Chionochloa rigida. N.Z. Journal of Botany 3: 300-19.

Martin, P. S.; Gray, J. 1962: Pollen analysis and the Cainozoic. Science 137: 103-11.

Mildenhall, D. S. 1972: Fossil pollen of Acacia type from New Zealand. N.Z. Journal of Botany 10: 485-94.

—— 1976: Exotic pollen rain on the Chatham Islands during the late Pleistocene. N.Z. Journal of Geology and Geophysics 19:327-33.

—— 1980: New Zealand late Cretaceous and Cenozoic plant biography; a contribution, Paleogeography, Paleoclimatology, paleoecology 31: 197-233.

Moar, N. T. 1969: Pollen analysis of a surface sample from Antipodes Island. N.Z. Journal of Botany 7: 419-23.

—— 1970: Recent pollen spectra from three localities in the South Island, New Zealand. N.Z. Journal of Botany 8: 210-21.

—— 1971: Contributions to the Quaternary history of the New Zealand flora 6. Aranuian pollen diagrams from Canterbury, Nelson, and North Westland, South Island. N.Z. Journal of Botany 9: 80-145.

—— 1973: Late Pleistocene vegetation and environment in southern New Zealand. Paleoecology of Africa 8: 181-98.

Moar, N. T.; Suggate, R. P. 1979: Contributions to the Quaternary history of the New Zealand flora 8. Inerglacial and glacial vegetation in the Westport district, South Island. N.Z. Journal of Botany 17: 316-87.

Molloy, B. P. 1969: Evidence for post-glacial climatic change in New Zealand, Journal of Hydrology (N.Z.) 8: 56-67.

Molloy, B. P.; Cox, J. E. 1972: Subfossil forest remains and their bearing on forest history in the Rakaia Catchment, Canterbury, New Zealand. N.Z. Journal of Botany 10: 267-76.

Moore, L. B.; Edgar, E. 1970: “Flora of New Zealand Vol. 2. Government Printer, Wellington 345pp.

Myers, V. V. 1973: A note on the dispersal of Nothofagus pollen in Canterbury. N.Z. Journal of Botany 11: 311-16.

Nicholls, J. L. 1976: A revised classification of the North Island indigenous forests. N.Z. Journal of Forestry 21: 105-32.

Oldfield, F. 1970: Some aspects of scale and complexity in pollen-analytically based paleoecology. Pollen et Spores 12: 163-71.

Park, G. N.; Walls, G. Y. 1978: Inventory of tall forest stands on plains and terraces in the Nelson and Marlborough Land Districts, New Zealand. Unpubl. report. Botany Division, D.S.I.R.

Pocknall, D. T. 1978: Relative pollen representation in relation to vegetation composition, Westland, New Zealand. N.Z. Journal of Botany 16: 379-86.

—— 1979: “Studies on the Pollen of the New Zealand Gymnosperms . Unpubl. Ph.D. thesis, University of Canterbury, Christchurch.

—— 1980: Modern pollen rain and Aranuian vegetation from Lady Lake, northwestland, New Zealand. N.Z. Journal of Botany 18: 275-84.

Poole, A. L. 1937: A brief ecological survey of Mt Pukekura State Forest, South Westland. N.Z. Journal of Forestry 4: 78-85.

Rattenbury, J. A. 1962: Cyclic hybridization as a survival mechanism in the New Zealand forest flora. Evolution 16: 348-63.

Rigg, H. M. 1962: The pakihi bogs of Westport, New Zealand. Transactions of the Royal Society of New Zealand, Botany, 1,7: 91-108.

Robbins, R. G. 1962: The podocarp-broadleaf forests of New Zealand. Transactions of the Royal Society of New Zealand (Botany) 1: 33-75.

Rymer, L. 1976: The late-Devensian flora of Drimgall, North Knapdale, Agryllshire. Transactions of the Botanical Society of Edinburgh 43: 55-65.

Sakai, A.; Paton, D. M.; Wardle, P. 1981: Freezing resistance of trees of the south temperate zone, especially subalpine species of Australasia. Ecology 62(3): 563-570.

Salgado-Labouriau, M. L. 1979: Modern pollen deposition in the Venezuelan Andes. Grana 18: 53-68.

Soons, J. M.; Burrows, C. J. 1978: Dates for Otiran deposits including plant microfossils and macrofossils from Rakaia Valley. N.Z. Journal of Geology and Geophysics 21: 607-15.

Specht, R. L. 1970: Vegetation. In “The Australian Environment G. W. Leeper, Ed.) Melbourne University Press, Melbourne pp. 45-67.

Stevenson, G. 1978: Botanical evidence linking the New Zealand Maoris with New Caledonia and the New Hebrides. Nature 276: 704-5.

Tauber, H. 1965: Differential pollen dispersal and the interpretation of pollen diagrams, Denmarks geologiske Undersogelse Series 11: 1-69.

—— 1967: Investigations into the mode of pollen tranfer in forested areas. Review of Paleobotany and Palynology 3: 277-87.

Walker, D.; Flenley, J. R. 1979: Late Quaternary vegetational history of the Enga Province of upland Papua New Guinea. Philosphical Transactions of the Royal Society of London B 286:265-344.

Wardle, J. 1970: The ecology of Nothofagus solandri 1-4 N.Z. Journal of Botany 8: 494-646.

Wardle, P. 1963: The regeneration gap of New Zealand gymnosperms N.Z. Journal of Botany 1: 301-315.

—— 1964: Facets of the distribution of forest vegetation in New Zealand. N.Z. Journal of Botany 2: 352-66.

—— 1966: Biological flora of New Zealand 1. Weinmannia racemosa Linn. f. (Cunoniaceae) Kamahi. N.Z. Journal of Botany 4:114-31.

—— 1967: Biological flora of New Zealand 2. Nothofagus menziesii (Hook f.) Oerst. (Fagaceae) Silver Beech. N.Z. Journal of Botany 5: 276-302.

—— 1969: Biological flora of New Zealand 4. Phyllocladus alpinus Hook f. (Podocarpaceae) Mountain Toatoa, Celery pine. N.Z. Journal of Botany 7: 76-95.

—— 1971a.: Effectiveness of distant pollen sources in Coprosma. N.Z. Journal of Botany 9: 223.

—— 1971b.: Biological flora of New Zealand 6. Metrosideros umbellata Cav. (Syn. M. lucida (Forst. f.) (A. Rich.) (Myrtaceae) Southern Rata. N.Z. Journal of Botany 9: 645-71.

—— 1973: Native vegetation, In “Natural History of New Zealand (G. R. Williams, Ed.) Reeds, Wellington: 155-169.

—— 1976: Vascular plants of the Westland National Park (New Zealand) and neighbouring lowland and coastal areas. N.Z. Journal of Botany 13: 497-454.

—— 1977: Plant communities of Westland National Park (New Zealand) and neighbouring lowland and coastal areas. N.Z. Journal of Botany 14: 323-98.

—— 1978: Regeneration status of some New Zealand conifers with particular reference to Libocedrus bidwillii in Westland National Park. N.Z. Journal of Botany 16:471-7.

—— 1980: Primary succession in Westland National Park and its vicinity, New Zealand. N.Z. Journal of Botany 18: 221-32.

Wardle, P; Campbell, A. D. 1976: Seasonal cycle of tolerance of low temperatures in three native woody plants, in relation to their ecology and post-glacial history. Proceedings of the New Zealand Ecological Society 23: 85-91.

Wardle, P.; T.R.O.; Spain, A. V. 1971: Biological flora of New Zealand 5. Olearia colensoi Hook f. (Compositae) Leatherwood, Tupari. N.Z. Journal of Botany 9: 186-214.

Wardle, P.; McKellar, M. H. 1978: Nothofagus menz/esrileaves dated at 7 490 yr B.P. in till-like sediments at Milford Sound, New Zealand. N.Z. Journal of Botany 16: 153-7.

Webb, L. J.; Tracy, J. G.; Williams, W. T. 1972: Regeneration and pattern in the subtropical rainforest. Journal of Ecolgy 60: 675-95.

Wright, H. E. 1967: The use of surface samples in Quaternary pollen analysis. Review of Paleobotany and Palynology 2: 231-30.